4. Occupation Trends of Top Occupations

Objective: To visualise how the occupations have changed across the three time periods of 1850, 1880 and 1910.

The graphing process showed that there were some inconsistencies with the coding of `occstr`, such as "DRESSMAKER" and "DRESS MAKER" which need to be addressed.

OCC1950 was incorporated to compare a coded variable consistent across all three census years.

1. Set Up

README

1850

library(dplyr)
library(ggplot2)
library(treemap)
library(extrafont)
library(stringr)
loadfonts()
library(knitr)
library(kableExtra)
library(tidyr)

# Loading
mn_1850 = readr::read_csv("../Data/census_1850_occ_mn.csv") %>% mutate(city = "Manhattan")
bk_1850 = readr::read_csv("../Data/census_1850_occ_bk.csv") %>% mutate(city = "Brooklyn")
OCC_1880 = readr::read_csv("../Data/OCC1880.csv")
OCC_1950 = readr::read_csv("../Data/OCC1950.csv")

# Combining Data
combined_1850 = rbind(mn_1850, bk_1850) %>% 
  select(age, sex, marst, race, labforce,
         occ, city, everything()) %>%
  left_join(OCC_1880, by = c("occ" = "code")) %>%
  mutate(race = factor(race,
                       levels = c(100,120,200,210,300),
                       labels = c("White", "Black (white)",
                                  "Black / African American",
                                  "Mulatto", "American Indian")),
         labforce = factor(labforce,
                           levels = c(0,1,2),
                           labels = c("N/A",
                                      "No, not in the labor force",
                                      "Yes, in the labor force")),
         sex = factor(sex,
                      levels = c(1,2),
                      labels = c("Male", "Female")),
         marst = factor(marst,
                        levels = c(1:6),
                        labels = c("Married, spouse present",
                                   "Married, spouse absent",
                                   "Separated",
                                   "Divorced",
                                   "Widowed",
                                   "Never married/single")))

# Theme setting for visualisations
theme_set(theme_minimal() + 
            theme(panel.grid.minor = element_blank(),
                  text = element_text(family = "Arial",
                                      colour = "black")))
col_sex = c("#e21737", "#0b648f")
col_miss = c("#86b817", "#e53238")

# Saving Memory
rm(OCC_1880)
rm(bk_1850)
rm(mn_1850)

1880

# Loading
mn_1880 = readr::read_csv("../Data/census_1880_occ_mn.csv") %>% mutate(city = "Manhattan")
bk_1880 = readr::read_csv("../Data/census_1880_occ_bk.csv") %>% mutate(city = "Brooklyn")
OCC_1880 = readr::read_csv("../Data/OCC1880.csv")

# Combining Data
combined_1880 = rbind(mn_1880, bk_1880) %>% 
  select(age, sex, race, labforce,
         occ, city, everything()) %>%
  left_join(OCC_1880, by = c("occ" = "code")) %>%
  mutate(age = ifelse(age == "Less than 1 year old",
                      0,
                      age),
         age = as.numeric(age))

rm(OCC_1880)
rm(bk_1880)
rm(mn_1880)

1910

# Loading
mn_1910 = readr::read_csv("../Data/census_1910_occ_mn.csv") %>% mutate(city = "Manhattan")
bk_1910 = readr::read_csv("../Data/census_1910_occ_bk.csv") %>% mutate(city = "Brooklyn")
OCC_1950 = readr::read_csv("../Data/OCC1950.csv")

# Combining Data
combined_1910 = rbind(mn_1910, bk_1910) %>%
  left_join(OCC_1950, by = c("occ1950" = "code")) %>%
  mutate(sex = factor(sex,
                      levels = c(1,2),
                      labels = c("Male", "Female")),
         labor_force = factor(labor_force,
                              levels = c(0,1,2),
                              labels = c("N/A", 
                                         "No, not in the labor force",
                                         "Yes, in the labor force")))

rm(OCC_1950)
rm(mn_1910)
rm(bk_1910)

2. Analysis of Occupation Trends

README

This section investigated how occupations have changed across the three time periods of 1850, 1880 and 1910. To this end, occstr was preferred over label due to the consistent categorising of occupations.

Again, as with the previous section of Occupation counts by year, the only respondents who were in the “Yes, in the labor force” level of labforce were included.

Specifically, the following tabs investigate the top 10 occupations of each time period, and a list of them is shown in the next tab.

Top 10 Jobs of each year, across time

The tabe below shows the top 10 jobs in each time period

top10_1850 = combined_1850 %>%
  filter(labforce == "Yes, in the labor force") %>%
  count(occstr) %>%
  arrange(desc(n)) %>%
  .[1:10, "occstr"]

top10_1880 = combined_1880 %>%
  filter(labforce == "Yes, in the labor force") %>%
  count(occstr) %>%
  arrange(desc(n)) %>%
  .[1:10, "occstr"]

top10_1910 = combined_1910 %>%
  filter(labor_force == "Yes, in the labor force" & 
           !is.na(occstr)) %>%
  count(occstr) %>%
  arrange(desc(n)) %>%
  .[1:10, "occstr"]

cbind(top10_1850, top10_1880, top10_1910) %>% 
  kable(col.names = c("1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

1850	1880	1910
LABORER	LABORER	LABORER
CLERK	SERVANT	CLERK
TAILOR	CLERK IN STORE	OPERATOR
CARPENTER	TAILOR	SALESMAN
MERCHANT	CARPENTER	SERVANT
SEAMAN	CLERK	TAILOR
GROCER	DRESS MAKER	DRIVER
MASON	BUTCHER	DRESSMAKER
SHOEMAKER	PAINTER	CARPENTER
CARTMAN	DRESSMAKER	STENOGRAPHER

• From the table of top 10 occupations, we can see that aside from the top jobs being laborers and clerks, there is an inconsistency with the coding of occstrin 1880, such as “DRESSMAKER” and “DRESS MAKER”.

• Subsequently, the code was re-ran, to count Dress Maker and Dressmaker together, along with Clerk and Clerk in store. On this run, further ammendments were added, such as to add Domestic Servant and Servant together. The code is not shown, but the new results are:

1850	1880	1910
LABORER	SERVANT	LABORER
CLERK	LABORER	CLERK
TAILOR	CLERK	OPERATOR
CARPENTER	DRESSMAKER	SALESMAN
MERCHANT	TAILOR	SERVANT
SEAMAN	CARPENTER	TAILOR
GROCER	BUTCHER	DRIVER
MASON	PAINTER	DRESSMAKER
SHOEMAKER	PRINTER	CARPENTER
CARTMAN	CIGAR MAKER	STENOGRAPHER

Table of Top Occupations

top10_combined = top10_years %>%
  left_join(combined_1850 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr),
            by = "occstr") %>%
  left_join(combined_1880 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr),
            by = "occstr") %>% 
  left_join(combined_1910 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr),
            by = "occstr") %>%
  select(occupation = occstr,
         t_1850 = n.x,
         t_1880 = n.y,
         t_1910 = n) %>%
  gather(key = "year", value = "count", -occupation) %>%
  mutate(year = case_when(year == "t_1850" ~ 1850,
                          year == "t_1880" ~ 1880,
                          year == "t_1910" ~ 1910),
         count = ifelse(is.na(count),
                        0,
                        count))

# Adding logical of if occupation was a top 10 for that particular year
top10_by_year = rbind(cbind(top10_1850, year = 1850),
        cbind(top10_1880, year = 1880),
        cbind(top10_1910, year = 1910)) %>%
  select(occupation = occstr, year) %>%
  mutate(top = 1)

top10_combined = top10_combined %>% 
  left_join(top10_by_year, by = c("occupation", "year")) %>% mutate(top = ifelse(is.na(top),
                                                                                 0,
                                                                                 1))

top10_combined %>% 
  select(occupation, year, top) %>% 
  spread(key = year, value = top)%>% 
  kable(col.names = c("occupation","1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

occupation	1850	1880	1910
BUTCHER	0	1	0
CARPENTER	1	1	1
CARTMAN	1	0	0
CIGAR MAKER	0	1	0
CLERK	1	1	1
DRESSMAKER	0	1	1
DRIVER	0	0	1
GROCER	1	0	0
LABORER	1	1	1
MASON	1	0	0
MERCHANT	1	0	0
OPERATOR	0	0	1
PAINTER	0	1	0
PRINTER	0	1	0
SALESMAN	0	0	1
SEAMAN	1	0	0
SERVANT	0	1	1
SHOEMAKER	1	0	0
STENOGRAPHER	0	0	1
TAILOR	1	1	1

• The table above indicates 1 if the occupation was a top 10 occupation in that time period, and 0 otherwise.
• It shows that some occupations became only popular in later time periods, such as “DRIVER”, while other occupations were only popular in earlier time periods such as “GROCER”.

Changes in Top 10 Occupations (Dot Plot)

col_3 = c("#ffa600", "#bc5090", "#003f5c")
top10_combined %>%
  ggplot(aes(y = reorder(occupation, count), 
             x = count, 
             col = factor(year),
             group = occupation)) + 
  geom_point(aes(size = count),
             alpha = 0.8,
             show.legend = FALSE) +
  geom_line() +
  scale_color_manual(values = col_3) +
  scale_radius() +
  scale_x_continuous(breaks = seq(0, 100000, 20000)) +
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Size of points vary on count") +
  theme(legend.position = "top")

• From the graph, we can see that laborer,and clerks were top occupations in 1910.
• Some occupations were top 10, but saw little change such as grocers, masons and cartmen.
• This also highlights a need to check for the coding of “Clerk” in 1880, as it might be a mistake in the categorising of a clerk, given that the count had dropped in 1880l, but increased greatly in 1910.

Changes in Top 10 Occupations (Line Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_point(aes(col = colour,
                 shape = factor(top)),
             size = 2) +
  geom_line(aes(col = colour)) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  scale_colour_manual(values = c("#ee4c58", "#56c1ab")) +
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Triangle points were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") + 
  theme(legend.position = "none")

Changes in Top 10 Occupations (Bar Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_col(aes(fill = colour,
               col = factor(top)),
           size = 1) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  
  scale_color_manual(values = c("white", "black")) +
  scale_fill_manual(values = c("#ee4c58", "#56c1ab")) +
  scale_x_continuous(breaks = c(1850, 1880, 1910)) +
  
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Black borders were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") +
  theme(legend.position = "none")

3. Analysis of Occupation Trends (%)

README

This section investigated how occupations have changed across the three time periods of 1850, 1880 and 1910. To this end, occstr was preferred over label due to the consistent categorising of occupations.

Again, as with the previous section of Occupation counts by year, the only respondents who were in the “Yes, in the labor force” level of labforce were included.

Specifically, the following tabs investigate the top 10 occupations of each time period, and a list of them is shown in the next tab.

Top 10 Jobs of each year, across time

The tabe below shows the top 10 jobs in each time period

top10_1850 = combined_1850 %>%
  filter(labforce == "Yes, in the labor force") %>%
  count(occstr) %>%
  mutate(perc_lf = round(n / sum(combined_1850$labforce == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1850) * 100, 
                          digits = 2)) %>% 
  arrange(desc(n)) %>%
  .[1:10, "occstr"]

top10_1880 = combined_1880 %>%
  filter(labforce == "Yes, in the labor force") %>%
  count(occstr) %>%
  mutate(perc_lf = round(n / sum(combined_1880$labforce == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1880) * 100, 
                          digits = 2)) %>%
  arrange(desc(n)) %>% 
    .[1:10, "occstr"]

top10_1910 = combined_1910 %>%
  filter(labor_force == "Yes, in the labor force" & 
           !is.na(occstr)) %>%
  count(occstr) %>%
  mutate(perc_lf = round(n / sum(combined_1910$labor_force == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1910) * 100, 
                          digits = 2)) %>%
  arrange(desc(n)) %>%
  .[1:10, "occstr"]

cbind(top10_1850, top10_1880, top10_1910) %>% 
  kable(col.names = c("1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

1850	1880	1910
LABORER	LABORER	LABORER
CLERK	SERVANT	CLERK
TAILOR	CLERK IN STORE	OPERATOR
CARPENTER	TAILOR	SALESMAN
MERCHANT	CARPENTER	SERVANT
SEAMAN	CLERK	TAILOR
GROCER	DRESS MAKER	DRIVER
MASON	BUTCHER	DRESSMAKER
SHOEMAKER	PAINTER	CARPENTER
CARTMAN	DRESSMAKER	STENOGRAPHER

• From the table of top 10 occupations, we can see that aside from the top jobs being laborers and clerks, there is an inconsistency with the coding of occstrin 1880, such as “DRESSMAKER” and “DRESS MAKER”.

• Subsequently, the code was re-ran, to count Dress Maker and Dressmaker together, along with Clerk and Clerk in store. On this run, further ammendments were added, such as to add Domestic Servant and Servant together. The code is not shown, but the new results are:

1850	1880	1910
LABORER	SERVANT	LABORER
CLERK	LABORER	CLERK
TAILOR	CLERK	OPERATOR
CARPENTER	DRESSMAKER	SALESMAN
MERCHANT	TAILOR	SERVANT
SEAMAN	CARPENTER	TAILOR
GROCER	BUTCHER	DRIVER
MASON	PAINTER	DRESSMAKER
SHOEMAKER	PRINTER	CARPENTER
CARTMAN	CIGAR MAKER	STENOGRAPHER

Table of Top Occupations

top10_combined = top10_years %>%
  left_join(combined_1850 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr) %>% 
                    mutate(perc_lf = round(n / sum(combined_1850$labforce == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1850) * 100, 
                            digits = 2)),
            by = "occstr") %>%
  left_join(combined_1880 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr) %>% 
                mutate(perc_lf = round(n / sum(combined_1880$labforce == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1880) * 100, 
                            digits = 2)),
            by = "occstr") %>% 
  left_join(combined_1910 %>%
              filter(occstr %in% top10_years$occstr) %>%
              count(occstr) %>% 
                mutate(perc_lf = round(n / sum(combined_1910$labor_force == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1910) * 100, 
                            digits = 2)),
            by = "occstr") %>%
  select(occupation = occstr,
         perc_lf_1850 = perc_lf.x,
         perc_lf_1880 = perc_lf.y,
         perc_lf_1910 = perc_lf) %>%
  gather(key = "year", value = "count", -occupation) %>%
  mutate(year = case_when(year == "perc_lf_1850" ~ 1850,
                          year == "perc_lf_1880" ~ 1880,
                          year == "perc_lf_1910" ~ 1910),
         count = ifelse(is.na(count),
                        0,
                        count))

# Adding logical of if occupation was a top 10 for that particular year
top10_by_year = rbind(cbind(top10_1850, year = 1850),
        cbind(top10_1880, year = 1880),
        cbind(top10_1910, year = 1910)) %>%
  select(occupation = occstr, year) %>%
  mutate(top = 1)

top10_combined = top10_combined %>% 
  left_join(top10_by_year, by = c("occupation", "year")) %>% mutate(top = ifelse(is.na(top),
                                                                                 0,
                                                                                 1))

top10_combined %>% 
  select(occupation, year, top) %>% 
  spread(key = year, value = top)%>% 
  kable(col.names = c("occupation","1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

occupation	1850	1880	1910
BUTCHER	0	1	0
CARPENTER	1	1	1
CARTMAN	1	0	0
CIGAR MAKER	0	1	0
CLERK	1	1	1
DRESSMAKER	0	1	1
DRIVER	0	0	1
GROCER	1	0	0
LABORER	1	1	1
MASON	1	0	0
MERCHANT	1	0	0
OPERATOR	0	0	1
PAINTER	0	1	0
PRINTER	0	1	0
SALESMAN	0	0	1
SEAMAN	1	0	0
SERVANT	0	1	1
SHOEMAKER	1	0	0
STENOGRAPHER	0	0	1
TAILOR	1	1	1

• The table above indicates 1 if the occupation was a top 10 occupation in that time period, and 0 otherwise.
• It shows that some occupations became only popular in later time periods, such as “DRIVER”, while other occupations were only popular in earlier time periods such as “GROCER”.

Changes in Top 10 Occupations (Dot Plot)

col_3 = c("#ffa600", "#bc5090", "#003f5c")
top10_combined %>%
  ggplot(aes(y = reorder(occupation, count), 
             x = count, 
             col = factor(year),
             group = occupation)) + 
  geom_point(aes(size = count),
             alpha = 0.8,
             show.legend = FALSE) +
  geom_line() +
  scale_color_manual(values = col_3) +
  scale_radius() +
  scale_x_continuous() +
  labs(y = "Occupations", col = "Year", x = "Percent of Labor Force",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Size of points vary on count") +
  theme(legend.position = "top")

• From the graph, we can see that laborer,and clerks were top occupations in 1910.
• Some occupations were top 10, but saw little change such as grocers, masons and cartmen.
• This also highlights a need to check for the coding of “Clerk” in 1880, as it might be a mistake in the categorising of a clerk, given that the count had dropped in 1880l, but increased greatly in 1910.

Changes in Top 10 Occupations (Line Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_point(aes(col = colour,
                 shape = factor(top)),
             size = 2) +
  geom_line(aes(col = colour)) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  scale_colour_manual(values = c("#ee4c58", "#56c1ab")) +
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Percentages in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Triangle points were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") + 
  theme(legend.position = "none")

Changes in Top 10 Occupations (Bar Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_col(aes(fill = colour,
               col = factor(top)),
           size = 1) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  
  scale_color_manual(values = c("white", "black")) +
  scale_fill_manual(values = c("#ee4c58", "#56c1ab")) +
  scale_x_continuous(breaks = c(1850, 1880, 1910)) +
  
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Percentages in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Black borders were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") +
  theme(legend.position = "none")

4. Analysis of Occupation Trends (OCC1950)

README

This section investigated how occupations have changed across the three time periods of 1850, 1880 and 1910. OCC1950 was used in this section to contrast the use of occstr.

Again, as with the previous section of Occupation counts by year, the only respondents who were in the “Yes, in the labor force” level of labforce were included.

Specifically, the following tabs investigate the top 10 occupations of each time period, and a list of them is shown in the next tab.

Top 10 Jobs of each year, across time

The tabe below shows the top 10 jobs in each time period

OCC_1950 = readr::read_csv("../Data/OCC1950.csv")

## Parsed with column specification:
## cols(
##   code = col_double(),
##   label = col_character()
## )

top10_1850 = combined_1850 %>%
    left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
  filter(labforce == "Yes, in the labor force") %>%
  count(label.y) %>%
  arrange(desc(n)) %>%
    .[1:10, "label.y"]

top10_1880 = combined_1880 %>%
    left_join(OCC_1950, by = c("occ1950" = "label")) %>% 
  filter(labforce == "Yes, in the labor force") %>%
  count(occ1950) %>%
  arrange(desc(n)) %>%
  .[1:10, "occ1950"]

top10_1910 = combined_1910 %>%
    left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
  filter(labor_force == "Yes, in the labor force" & 
           !is.na(occ1950)) %>%
  count(label.y) %>%
  arrange(desc(n)) %>%
  .[1:10, "label.y"]

cbind(top10_1850, top10_1880, top10_1910) %>% 
  kable(col.names = c("1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

1850	1880	1910
Laborers (n.e.c.)	Operative and kindred workers (n.e.c.)	Not yet classified
Managers, officials, and proprietors (n.e.c.)	Managers, officials, and proprietors (n.e.c.)	Operative and kindred workers (n.e.c.)
Operative and kindred workers (n.e.c.)	Private household workers (n.e.c.)	Managers, officials, and proprietors (n.e.c.)
Salesmen and sales clerks (n.e.c.)	Laborers (n.e.c.)	Salesmen and sales clerks (n.e.c.)
Carpenters	Salesmen and sales clerks (n.e.c.)	Private household workers (n.e.c.)
Tailors and tailoresses	Dressmakers and seamstresses, except factory	Laborers (n.e.c.)
Sailors and deck hands	Tailors and tailoresses	Clerical and kindred workers (n.e.c.)
Brickmasons, stonemasons, and tile setters	Carpenters	Tailors and tailoresses
Truck and tractor drivers	Truck and tractor drivers	Dressmakers and seamstresses, except factory
Cabinetmakers	Clerical and kindred workers (n.e.c.)	Bookkeepers

1850	1880	1910
Laborers (n.e.c.)	Operative and kindred workers (n.e.c.)	Not yet classified
Managers, officials, and proprietors (n.e.c.)	Managers, officials, and proprietors (n.e.c.)	Operative and kindred workers (n.e.c.)
Operative and kindred workers (n.e.c.)	Private household workers (n.e.c.)	Managers, officials, and proprietors (n.e.c.)
Salesmen and sales clerks (n.e.c.)	Laborers (n.e.c.)	Salesmen and sales clerks (n.e.c.)
Carpenters	Salesmen and sales clerks (n.e.c.)	Private household workers (n.e.c.)
Tailors and tailoresses	Dressmakers and seamstresses, except factory	Laborers (n.e.c.)
Sailors and deck hands	Tailors and tailoresses	Clerical and kindred workers (n.e.c.)
Brickmasons, stonemasons, and tile setters	Carpenters	Tailors and tailoresses
Truck and tractor drivers	Truck and tractor drivers	Dressmakers and seamstresses, except factory
Cabinetmakers	Clerical and kindred workers (n.e.c.)	Bookkeepers

Table of Top Occupations

top10_combined = top10_years %>%
  left_join(combined_1850 %>%
              left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
              filter(label.y %in% top10_years$label.y) %>%
              count(label.y),
            by = "label.y") %>%
  left_join(combined_1880 %>%
              left_join(OCC_1950, by = c("occ1950" = "label")) %>% 
              filter(occ1950 %in% top10_years$label.y) %>%
              count(occ1950),
            by = c("label.y"="occ1950")) %>% 
  left_join(combined_1910 %>%
              left_join(OCC_1950, by = c("occ1950" = "code")) %>%
              filter(label.y %in% top10_years$label.y) %>%
              count(label.y),
            by = "label.y") %>%
  select(occupation = label.y,
         t_1850 = n.x,
         t_1880 = n.y,
         t_1910 = n) %>%
  gather(key = "year", value = "count", -occupation) %>%
  mutate(year = case_when(year == "t_1850" ~ 1850,
                          year == "t_1880" ~ 1880,
                          year == "t_1910" ~ 1910),
         count = ifelse(is.na(count),
                        0,
                        count))
# Adding logical of if occupation was a top 10 for that particular year
top10_by_year = rbind(cbind(top10_1850, year = 1850),
        cbind(top10_1880 %>% 
                  rename(label.y = occ1950), year = 1880),
        cbind(top10_1910, year = 1910)) %>%
  select(occupation = label.y, year) %>%
  mutate(top = 1)

top10_combined = top10_combined %>% 
  left_join(top10_by_year, by = c("occupation", "year")) %>% mutate(top = ifelse(is.na(top),
                                                                                 0,
                                                                                 1))

top10_combined %>% 
  select(occupation, year, top) %>% 
  spread(key = year, value = top)%>% 
  kable(col.names = c("occupation","1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

occupation	1850	1880	1910
Bookkeepers	0	0	1
Brickmasons, stonemasons, and tile setters	1	0	0
Cabinetmakers	1	0	0
Carpenters	1	1	0
Clerical and kindred workers (n.e.c.)	0	1	1
Dressmakers and seamstresses, except factory	0	1	1
Laborers (n.e.c.)	1	1	1
Managers, officials, and proprietors (n.e.c.)	1	1	1
Not yet classified	0	0	1
Operative and kindred workers (n.e.c.)	1	1	1
Private household workers (n.e.c.)	0	1	1
Sailors and deck hands	1	0	0
Salesmen and sales clerks (n.e.c.)	1	1	1
Tailors and tailoresses	1	1	1
Truck and tractor drivers	1	1	0

• The table above indicates 1 if the occupation was a top 10 occupation in that time period, and 0 otherwise.

Changes in Top 10 Occupations (Dot Plot)

col_3 = c("#ffa600", "#bc5090", "#003f5c")
top10_combined %>%
  ggplot(aes(y = reorder(occupation, count), 
             x = count, 
             col = factor(year),
             group = occupation)) + 
  geom_point(aes(size = count),
             alpha = 0.8,
             show.legend = FALSE) +
  geom_line() +
  scale_color_manual(values = col_3) +
  scale_radius() +
  scale_x_continuous(labels = scales::comma) +
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Size of points vary on count") +
  theme(legend.position = "top")

Changes in Top 10 Occupations (Line Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_point(aes(col = colour,
                 shape = factor(top)),
             size = 2) +
  geom_line(aes(col = colour)) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  scale_colour_manual(values = c("#ee4c58", "#56c1ab")) +
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Triangle points were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") + 
  theme(legend.position = "none")

Changes in Top 10 Occupations (Bar Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_col(aes(fill = colour,
               col = factor(top)),
           size = 1) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  
  scale_color_manual(values = c("white", "black")) +
  scale_fill_manual(values = c("#ee4c58", "#56c1ab")) +
  scale_x_continuous(breaks = c(1850, 1880, 1910)) +
  
  labs(y = "Occupations", col = "Year", x = "Count",
       title = "Graph of Occupation Counts in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Black borders were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") +
  theme(legend.position = "none")

5. Analysis of Occupation Trends (OCC1950) (%)

README

This section investigated how occupations have changed across the three time periods of 1850, 1880 and 1910. OCC1950 was used in this section to contrast the use of occstr.

Again, as with the previous section of Occupation counts by year, the only respondents who were in the “Yes, in the labor force” level of labforce were included.

Specifically, the following tabs investigate the top 10 occupations of each time period, and a list of them is shown in the next tab.

Top 10 Jobs of each year, across time

The tabe below shows the top 10 jobs in each time period

OCC_1950 = readr::read_csv("../Data/OCC1950.csv")

## Parsed with column specification:
## cols(
##   code = col_double(),
##   label = col_character()
## )

top10_1850 = combined_1850 %>%
    left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
  filter(labforce == "Yes, in the labor force") %>%
  count(label.y) %>%
  mutate(perc_lf = round(n / sum(combined_1850$labforce == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1850) * 100, 
                          digits = 2)) %>%
  arrange(desc(n)) %>%
    .[1:10, "label.y"]

top10_1880 = combined_1880 %>%
    left_join(OCC_1950, by = c("occ1950" = "label")) %>% 
  filter(labforce == "Yes, in the labor force") %>%
  count(occ1950) %>%
  mutate(perc_lf = round(n / sum(combined_1910$labor_force == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1910) * 100, 
                          digits = 2)) %>% 
  arrange(desc(n)) %>%
  .[1:10, "occ1950"]

top10_1910 = combined_1910 %>%
    left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
  filter(labor_force == "Yes, in the labor force" & 
           !is.na(occ1950)) %>%
  count(label.y) %>%
  mutate(perc_lf = round(n / sum(combined_1910$labor_force == "Yes, in the labor force") * 100,
                         digits = 2),
         perc_pop = round(n / NROW(combined_1910) * 100, 
                          digits = 2)) %>%
  arrange(desc(n)) %>%
  .[1:10, "label.y"]

cbind(top10_1850, top10_1880, top10_1910) %>% 
  kable(col.names = c("1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

1850	1880	1910
Laborers (n.e.c.)	Operative and kindred workers (n.e.c.)	Not yet classified
Managers, officials, and proprietors (n.e.c.)	Managers, officials, and proprietors (n.e.c.)	Operative and kindred workers (n.e.c.)
Operative and kindred workers (n.e.c.)	Private household workers (n.e.c.)	Managers, officials, and proprietors (n.e.c.)
Salesmen and sales clerks (n.e.c.)	Laborers (n.e.c.)	Salesmen and sales clerks (n.e.c.)
Carpenters	Salesmen and sales clerks (n.e.c.)	Private household workers (n.e.c.)
Tailors and tailoresses	Dressmakers and seamstresses, except factory	Laborers (n.e.c.)
Sailors and deck hands	Tailors and tailoresses	Clerical and kindred workers (n.e.c.)
Brickmasons, stonemasons, and tile setters	Carpenters	Tailors and tailoresses
Truck and tractor drivers	Truck and tractor drivers	Dressmakers and seamstresses, except factory
Cabinetmakers	Clerical and kindred workers (n.e.c.)	Bookkeepers

1850	1880	1910
Laborers (n.e.c.)	Operative and kindred workers (n.e.c.)	Not yet classified
Managers, officials, and proprietors (n.e.c.)	Managers, officials, and proprietors (n.e.c.)	Operative and kindred workers (n.e.c.)
Operative and kindred workers (n.e.c.)	Private household workers (n.e.c.)	Managers, officials, and proprietors (n.e.c.)
Salesmen and sales clerks (n.e.c.)	Laborers (n.e.c.)	Salesmen and sales clerks (n.e.c.)
Carpenters	Salesmen and sales clerks (n.e.c.)	Private household workers (n.e.c.)
Tailors and tailoresses	Dressmakers and seamstresses, except factory	Laborers (n.e.c.)
Sailors and deck hands	Tailors and tailoresses	Clerical and kindred workers (n.e.c.)
Brickmasons, stonemasons, and tile setters	Carpenters	Tailors and tailoresses
Truck and tractor drivers	Truck and tractor drivers	Dressmakers and seamstresses, except factory
Cabinetmakers	Clerical and kindred workers (n.e.c.)	Bookkeepers

Table of Top Occupations

top10_combined = top10_years %>%
  left_join(combined_1850 %>%
              left_join(OCC_1950, by = c("occ1950" = "code")) %>% 
              filter(label.y %in% top10_years$label.y) %>%
              count(label.y) %>% 
              mutate(perc_lf = round(n / sum(combined_1850$labforce == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1850) * 100, 
                            digits = 2)),
            by = "label.y") %>%
  left_join(combined_1880 %>%
              left_join(OCC_1950, by = c("occ1950" = "label")) %>% 
              filter(occ1950 %in% top10_years$label.y) %>%
              count(occ1950) %>% 
              mutate(perc_lf = round(n / sum(combined_1880$labforce == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1880) * 100, 
                            digits = 2)),
            by = c("label.y"="occ1950")) %>% 
  left_join(combined_1910 %>%
              left_join(OCC_1950, by = c("occ1950" = "code")) %>%
              filter(label.y %in% top10_years$label.y) %>%
              count(label.y) %>% 
              mutate(perc_lf = round(n / sum(combined_1910$labor_force == "Yes, in the labor force") * 100,
                           digits = 2),
           perc_pop = round(n / NROW(combined_1910) * 100, 
                            digits = 2)),
            by = "label.y") %>%
  select(occupation = label.y,
         perc_lf_1850 = perc_lf.x,
         perc_lf_1880 = perc_lf.y,
         perc_lf_1910 = perc_lf) %>%
  gather(key = "year", value = "count", -occupation) %>%
  mutate(year = case_when(year == "perc_lf_1850" ~ 1850,
                          year == "perc_lf_1880" ~ 1880,
                          year == "perc_lf_1910" ~ 1910),
         count = ifelse(is.na(count),
                        0,
                        count))
# Adding logical of if occupation was a top 10 for that particular year
top10_by_year = rbind(cbind(top10_1850, year = 1850),
        cbind(top10_1880 %>% 
                  rename(label.y = occ1950), year = 1880),
        cbind(top10_1910, year = 1910)) %>%
  select(occupation = label.y, year) %>%
  mutate(top = 1)

top10_combined = top10_combined %>% 
  left_join(top10_by_year, by = c("occupation", "year")) %>% mutate(top = ifelse(is.na(top),
                                                                                 0,
                                                                                 1))

top10_combined %>% 
  select(occupation, year, top) %>% 
  spread(key = year, value = top)%>% 
  kable(col.names = c("occupation","1850", "1880", "1910")) %>%
  kable_styling() %>%
  scroll_box(width = "100%", height = "300px")

occupation	1850	1880	1910
Bookkeepers	0	0	1
Brickmasons, stonemasons, and tile setters	1	0	0
Cabinetmakers	1	0	0
Carpenters	1	1	0
Clerical and kindred workers (n.e.c.)	0	1	1
Dressmakers and seamstresses, except factory	0	1	1
Laborers (n.e.c.)	1	1	1
Managers, officials, and proprietors (n.e.c.)	1	1	1
Not yet classified	0	0	1
Operative and kindred workers (n.e.c.)	1	1	1
Private household workers (n.e.c.)	0	1	1
Sailors and deck hands	1	0	0
Salesmen and sales clerks (n.e.c.)	1	1	1
Tailors and tailoresses	1	1	1
Truck and tractor drivers	1	1	0

• The table above indicates 1 if the occupation was a top 10 occupation in that time period, and 0 otherwise.

Changes in Top 10 Occupations (Dot Plot)

col_3 = c("#ffa600", "#bc5090", "#003f5c")
top10_combined %>%
  ggplot(aes(y = reorder(occupation, count), 
             x = count, 
             col = factor(year),
             group = occupation)) + 
  geom_point(aes(size = count),
             alpha = 0.8,
             show.legend = FALSE) +
  geom_line() +
  scale_color_manual(values = col_3) +
  scale_radius() +
  scale_x_continuous(labels = scales::comma) +
  labs(y = "Occupations", col = "Year", x = "Percent",
       title = "Graph of Occupation % in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Size of points vary on count") +
  theme(legend.position = "top")

Changes in Top 10 Occupations (Line Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_point(aes(col = colour,
                 shape = factor(top)),
             size = 2) +
  geom_line(aes(col = colour)) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  scale_colour_manual(values = c("#ee4c58", "#56c1ab")) +
  labs(y = "Occupations", col = "Year", x = "Percent",
       title = "Graph of Occupation % in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Triangle points were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") + 
  theme(legend.position = "none")

Changes in Top 10 Occupations (Bar Graph)

top10_combined %>%
  group_by(occupation) %>%
  mutate(difference = first(count) - last(count),
         colour = ifelse(difference > 0,
                         "#86b817",
                         "#e53238")) %>%
  ggplot(aes(y = count, x = year)) +
  geom_col(aes(fill = colour,
               col = factor(top)),
           size = 1) +
  facet_wrap(~reorder(occupation, difference),
             scales = "free") +
  
  scale_color_manual(values = c("white", "black")) +
  scale_fill_manual(values = c("#ee4c58", "#56c1ab")) +
  scale_x_continuous(breaks = c(1850, 1880, 1910)) +
  
  labs(y = "Occupations", col = "Year", x = "Percent",
       title = "Graph of Occupation % in Top 10 Occupations from 1850, 1880 and 1910",
       caption = "Note: Black borders were if the occupation was a top 10 occupation in that time\nperiod. Occupations areordered by the size of change from 1850 to 1910") +
  theme(legend.position = "none")